Optionality Has No Router

I. The Trace

A positive-sum interaction may exist. The artifact exists. The relevant evaluator exists. The sender is willing to bound the ask. The recipient might benefit. But every routing layer fails.

Direct message: unavailable. Public mention: invisible. Forum post: zero engagement. Email: filtered or bounced. Institutional intake: absent. Search: useless, because nobody types the query that would find it. The sender has no institutional affiliation to collapse the verification cost. The recipient has no declared policy for what kinds of signals they accept. No intermediary exists who could vouch for the match. The possibility dies in the gap between “this exists” and “the right person sees it.”

This is the default state of unaffiliated optionality. Every attempt to make pay-to-read into a general protocol either failed or pivoted — the $5–50 price range creates adverse selection: too expensive for newcomers, too cheap for anyone whose time is valuable. The market tried. The gap remains.

Research on cold outreach shows that senders using custom business domains achieve roughly double the reply rate of freemail accounts — a penalty enforced algorithmically before any human sees the message. In academia, faculty ignore requests from women and minorities at measurably higher rates, especially in higher-paying disciplines. The institutional filter correlates with quality on average. It also systematically excludes the cross-domain, early-stage, and structurally novel signals that no institution produces internally.

II. The Four Existing Routers

Civilization currently routes signals through four mechanisms. Each works. Each fails in the same place.

Money routes by willingness to pay. It filters cheap talk effectively. It also creates plutocratic access, adverse selection (the people willing to pay for attention are disproportionately salespeople), and a social norm violation that destroys trust in contexts where communication should be free. Every pay-to-message startup discovered that attaching a price to human interaction triggers rejection, not engagement. Money is useful as one form of stake or rate-limiting. It fails when it becomes the sole routing mechanism.

Institutions route by affiliation. A university domain, a company name, a job title — these collapse verification cost to near zero. The institution has already vetted the sender. But institutional routing suppresses exactly the signals institutions cannot produce: cross-domain synthesis, structural critique of the institution itself, and work from people whose career trajectory doesn’t fit any existing department. The engineering meta level is empty partly because no institution routes signals to it.

Social graphs route by proximity. Warm introductions work — a mutual connection transfers trust from the known to the unknown. But social graphs trap optionality inside existing networks. Attention concentrates on already-prominent nodes through preferential attachment. The people with the most valuable cross-domain signals are, by definition, outside every existing graph. They fell out of every pipeline. Nobody in the graph knows them.

Platforms route by engagement or semantic similarity. Algorithms scale routing to billions of users. They also optimize for psychological capture — dwell time, click-through, completion rate — not for the actual value of the signal to the recipient. AI makes this worse: plausible-sounding relevant text is now infinitely cheap to generate. Content matching can identify that a message is topically relevant. It cannot assess whether the message is true, important, or worth interrupting someone.

The valuable cases that most need routing are early, illegible, cross-domain, socially distant, or low-status. Every existing router is weakest there.

III. The Missing Primitive

The missing piece is not better platforms, more money, or wider social networks. It is a primitive that none of the four routers implement:

An accountable routing claim. A claim of the form: “I believe this possible interaction is worth bounded attention from this evaluator class, and I accept consequences if that judgment is careless.”

The primitive is: make a bounded option legible, route it to the right class, attach accountable judgment, obtain consent where needed, collect outcome feedback.

Current systems route by sender identity: who are you, where do you work, who follows you. A better system routes by claim structure: what is the claim, who can evaluate it, what attention is needed, who has vouched for it, what are they risking.

“Please read my essay” is noise. “Three-minute triage request concerning a missing evaluation layer in legislative systems, supported by a working prototype, seeking classification as irrelevant or worth forwarding” — that can be handled by infrastructure.

IV. The Decision Rule

Route if the expected mutual value exceeds the attention cost, privacy cost, abuse risk, and routing cost combined. At the level of abstraction, this applies to research collaboration, governance correction, hiring, commerce, expert review, mentorship, and even some forms of social introduction. The implementation differs sharply by domain — hiring has legal constraints, romance requires mutual consent and easy refusal, scientific review needs evidence standards — but the shared primitive is the bounded, accountable claim that a possible interaction may be worth attention.

Dating apps route by attraction signals and engagement, not mutual flourishing. Hiring routes by credentials and keywords, not capability fit. Peer review routes by field and affiliation, not evaluator competence. Policy routes by lobbying access and salience, not expected public value. Each domain has its own broken router. Each would benefit from accountable, bounded optionality routing.

V. Moloch and the Missing Correction Path

The coordination failures described as “Moloch” — competitive pressures that produce collectively terrible outcomes — are often discussed as if they were a force. But a system can contain a corrective signal and still fail to correct. The signal may never reach the evaluator, may reach them in the wrong format, may arrive without credibility, may lack authority to act, or may impose too much attention cost. In those cases, Moloch is not the absence of knowledge. It is the absence of a correction path.

Cochrane’s Effectiveness and Efficiency appeared in the early 1970s; the Cochrane Collaboration was founded in 1993 to make systematic reviews into durable infrastructure. The correction signal existed for two decades before it became an institution with routing, maintenance, and authority. Deming taught Japanese industrial leaders in 1950; US manufacturers took his methods seriously in the 1980s, under competitive pressure from Japanese quality. The correction existed before the domestic routing became urgent.

Routing is not all of Moloch. Some coordination failures persist because of genuine incentive conflicts, enforcement gaps, or coalition structure. But routing is one of the missing organs. The engineering meta level produces correction signals. The mechanism authority acts on them. Between the two sits a gap: the routing layer that carries findings from diagnosis to action. Meta-level diagnosis without routing is Cochrane before the Collaboration. Routing without authority is inbox triage. Authority without meta-level diagnosis is bureaucracy. The full stack requires all three.

VI. The Universality

“Civilizational signal routing” sounds grand — as if only world-saving messages qualify. The opposite is true. The missing router handles any net-positive interaction that currently fails to happen because the parties cannot find each other, cannot verify each other, or cannot bound the cost of engagement.

The missed hire, the missed collaborator, the missed correction, the missed institution that should have formed but never did because the people who would have built it never met.

Civilization is the accumulation of routed positive-sum interactions. A civilization with better routing is one where more possibilities become actual — more Faradays find their Davys, more founders meet their co-founders, more policy failures reach someone who can act. The routing infrastructure is not a tool for civilization. It is civilization’s nervous system.

VII. What Fills It

Fragments of this already exist. Peer review uses editors as routers, manuscripts as packets, reviewers as endpoints, and editorial decisions as receipts. Expert networks sell routed access to expertise. Warm introductions stake social capital. Web-of-Trust systems verify identity. Decentralized identifiers standardize portable credentials. The novelty is narrow: separating the attention-worthiness claim from the truth claim and making that claim portable across domains. Most existing systems require “this sender is legitimate” and “this content is true” before they permit “this is worth your time.” The protocol makes the third claim independently routable.

Five objects. Each already exists informally. The protocol makes them explicit.

Optionality Packet. A bounded description of a possible positive-sum interaction. Not “please read my work” but a structured ask: what the claim is, who might care, what attention is requested, what evidence exists, what happens if it is ignored. The ask is bounded. Unbounded asks are expensive or unroutable.

Intent Endpoint. A declared policy for what kinds of options an actor is willing to inspect. Not an inbox — an explicit interface. “I accept three-minute triage requests about mechanism audit, institutional design, and AI governance. I do not accept debate requests or unbounded essays.” This turns attention from an unprotected commons into a specified protocol. Institutions and individuals publish what they review, what they ignore, and what format they require.

Router Attestation. A staked claim that a specific option is worth bounded attention from a specific evaluator class. The router endorses attention-worthiness, not truth. That distinction is the entire mechanism. Router incentives vary by domain: reputation in academic contexts, money in professional ones (sourcing, recruiting, and editorial triage are already paid routing labor), mandate in institutions (intake officers and grant managers are paid to classify inputs). The protocol makes the routing claim explicit so each domain can attach appropriate reward and penalty.

Actor Passport. Portable, domain-specific credibility independent of institutional affiliation. A universal reputation score becomes status again; domain-specific credibility (credible for Finnish law, credible for mechanism design, not yet credible for medicine) stays functional. A verifiable record of artifacts, routing outcomes, and attestations. Portable institutionality.

Outcome Receipt. Feedback that updates routing quality. A useful receipt can be one bit: relevant or irrelevant. It can be delegated to an agent. Recipients who return receipts get better future routing; endpoints that consume without feedback become sinks the network routes around. Without outcome feedback, the system degenerates into another inbox.

VIII. The Historical Pattern, Mechanized

Every historically successful outsider assembled enough of this stack by accident. Faraday bound his notes and hand-delivered them to Humphry Davy — an asymmetric proof packet routed to a specific evaluator. Ramanujan mailed his theorems to Hardy — a bounded ask with self-evident evidence. Each case required luck, class access, geographic accident, or geopolitical crisis.

The protocol mechanizes what they did informally. Asymmetric proof becomes the optionality packet. The insider champion becomes the accountable router. The warm introduction becomes the router attestation. The institutional title becomes the actor passport. Peer review becomes the outcome receipt.

The protocol reduces routing friction; it does not manufacture quality. Faraday still had to bind brilliant notes. When quality exists, routing should depend less on luck, class, proximity, and crisis.

IX. Spec as Common Knowledge

The first implementation is not software. The first implementation is making these objects socially real. Once people can say “that’s an optionality packet,” “that’s bad routing,” “that router has a track record,” “your ask is unbounded” — the infrastructure follows. A spec is not only an API. It is a social compression format. It tells everyone what kind of object they are looking at.

Agents reduce the cost of complying with the protocol — summarizing artifacts into packets, inferring recipient classes, tracking outcome receipts. But agents implement the protocol; they do not replace it. The spec must exist as common knowledge before agents can execute it, the same way HTTP had to exist before browsers could implement it.

Eight billion humans already broke email. A hundred billion agents will make trust-routing mandatory. The question is whether the protocol exists before that crisis or after.

X. The Bootstrap

The protocol begins as a small routing network in a high-value niche. One sender, one router, one recipient class, one outcome receipt. A narrow queue where structured packets outperform cold outreach, forum posts, and social media.

Bounded asks reduce unit cost but a thousand three-minute packets are still fifty hours. The system therefore needs layered triage. The first recipients are triage functions: editors, domain maintainers, community reviewers, institutional intake nodes, or trusted individuals willing to classify a narrow class of packets. Only after a packet survives triage does it escalate toward scarce expert attention. The path is earned, through prior outcomes, router endorsements, and compliance with declared endpoint policies — specific domains, attention budgets, and router trust thresholds.

XI. What This Looks Like

A public spec small enough to fit on one page.

Five objects: Packet, Intent, Attestation, Passport, Receipt. A valid route requires: bounded ask, declared target class, scope-limited endorsement, consent policy, outcome path. Everything else is optional infrastructure.

The first test: did the packet reach someone who could classify it better than the open internet did? Was the ask non-wasteful? Did the router’s judgment improve?

Civilization already routes optionality, badly, through institutions, money, social proximity, and virality. The question is whether positive-sum possibilities remain trapped inside those broken routers, or whether they get a protocol of their own.